CA2131433A1 - Mice deficient in neurotrophin receptors - Google Patents

Abstract

Abstract MICE DEFICIENT IN
NEUROTROPHIN RECEPTORS

The present invention provides mice and mouse cell lines having a homozygous or heterozygous deficiency in a gene encoding a neurotrophin receptor. In a preferred embodiment of this invention, mice and cell lines carry a trkB locus specifically targeted within its tyrosine protein kinase sequences. Mice homozygous for this mutation express gp95trkB receptor of unknown function but not the high affinity functional gp145trkB tyrosine protein kinase receptors. This mutation results in multiple CNS and PNS
neuronal deficiencies and in a postembryonic lethal phenotype.
Such genetically modified mice are useful in model systems for studying human diseases involving neuronal degeneration and neuronal cell loss, as well as in screening for genes, proteins, or other compounds that may prevent or impede neuronal cell death or stimulate neuronal regeneration.

Description

~ DC30 ,. ~ --1--, ' MICF DEFICIENT IN
NEUROTROPI IIN RECEPTORS

The present invention relates to the fields of neurology and genetically altered mice. Specifically, t~e pressnt invention relates to mice that are deficient in the ncrmal expression of one or more wild-type genes, to mica heterozygous for such a deliciency, and ~o cell lines that ar~ hsmozygous or heterozygous for this 1 0 deficiency.

The functional receptors responsible for mediating the trophic properties of the neNe grcYAh factor ~NGF) family ef neurotrophins have been recently idantified as membars of the ~.Ck family of tyrosine protein kinases (Barbaeid, 1993). Neuretrophins ar0 growth factors r~ ponsible for dsvelopment and maintenance of neurons. These mol~cules sxert their biological effect through high ~ffinity si~naling receptors located on the surface of spedfic typss o~ neurons.
To date, three different loci, designated ~, ~kB and I~C.
have been idantified. See U.S. Serial No. 837,814, filed February 25,1992, which is incorporated herein by refsrence. The product of the ~k proto~ncogene, a 140 IcDa cell surface tyrosine protein ~dnase designated as gpl40~k (Martin-Zanca Q~aL, 1989), is the high affinity recepior for NGF (Kaplan ~L. 1991; Klein ~L.
1991 a). The r01ated gpl 4~B tyrosine protBin kinase (Klein ~L. : :
l989~ selves as the si~naling recsptor ~ortwo related ~ ~ .

e 4 ~ ~ :

neurotrophins, brain-derived neurotrophic factor (BDNF) ~KIein ~L, 1991 b; Soppet ~L, 1991; Squinto ~L, 1991) and neurotrophin-4 (NT-4) (Barl<emeier~L, 1991, Klein t al., 1992; ~:
Ip ~" 1992). Finally, gp145~ , a ~yrosine protein kinase ancoded by the third member of this gene family, I~ , appears to ~ :
be primarily responsible for mediating the trophic properties of neurotrophin-3 (NT-3) (Lamballe .~1., 1991). This gene may also code tor additionai tyrosine protQin kinase isoforms that differ from gp14~kC by the presence of a small number of amino acid residues within their respective catalytic kinase domains (Lamballe ~L. submitted for publication).
The ~B gene is a large (>100 kbp) and compbx locus capable of direc~ing the synthesis of multipls transcripts (Klein ~L, 1989; 1990a; Middlemas *t al~, 1991). Some of these transcripts direct the synthe-~is of the gp145~ yrosine protein kinase recep~or. Other ~ranscripts however, code for a second class Of ~sB receptors that lack a catalytic kinase domain. Ons of these recæp10rs, ~pg5~kB, is abundantly expressed in adult mouse brain (Klein ~L 1990a). Nucleotide sequence analysis of cDNA
clones corr~sponding to these transcripts predicts that gpg5 has the same extracellular and transmembrane domains as gp1 45~kB, but contains a very short ~toplasmic r~gion o~ 23 amino acid r~idues of which tho last eleven bear no resemblance to any of the sequen~s presant in gp145~ Kbin ~t al.. 1 990a).
Mol0cular analysis of r~t h$B cDNA clones has identified a secor,d non catalytic ~kB r~c~ptor isoforrn with a predicl ed sequ6nce idertical ~o that of gpg511kB except ~or the presence of a unique nine amino aci~long sequance at its carboxyl tarminus (Middbmas ~LI 1991).
~ hybridizatien analysis has shown that the ~kB
gene is widely sxpressed in mu~ipls structur~s of the central and peripheral ne~vous systems (Klein ~, 1989; 1990a, b~. In the CNS, ~B transcripts have been observed in the cerebral cortex, hippocampus, thalamus, choroid plexus, granular layer of the '' ' : ~ ~' . , --~ - 3 -cerebellum, brain stern and spinal cord. In the PNS, ~CkB
exprassion is observed in many cranial ganglia including the trigeminal, facial, acoustic, IX/X superior and IX/X inferior ganglia, the retina and ophthalmic nerve, the vestibular system, multiple 5 facial structures, the subma~illary glands and dorsal root ganglia.
Analysis of ~B transcripts with probes specific for the catalytic and no n-catalytio ~kB receptors r~vsaled a rather dis~inct pattern of expression. For instance, in the adult mouse brain, gp145~kB
transcripts have been detected in the cerebral cortex, thalamus 10 and ~he pyramidal cell layer of the hippocampus. in oontrast, transcripts encoding the non catalytic gpg5~B receptor appear to be most prominent in structures containing non-neuronai cells such as the ependymal cell layer of the ventricles nd the choroid pl9XUs (Klein ~L, 1 990a).

Ths prssent invention provides mice and mouse cell lines that carry disrup~ed genes for eaoh of the ~k family of neurotrophin receptor gencs. This disruption, achieved by 20 homologous reoombination, completely abolishas expression of the corresponding high affinity signaling neurotrophin receptor.
Speoifically, the present invention describes mice wherein ths ~kB
Io~s has been modified so as to be deficisnt in exp~ession of ths wil~type ~kB gene product. In a preferred embodiment of lhis ~5 invention, mice carry a ~kB locus specHically targeted within its tyrosine protein hnase sequences. Mice homozygous for lhis mlnation axpress the non-catal~ic gpg~kB receptor of unknown function but not the high affinity h~nctional gp145~ tyrosir~e protein kinase roc~ptors. This muta~ion r0sults in mu~iple CNS :
30 ar;d PNS neuronal deficiencies and in a posgembryonic lethal phenotype.
Also in accordance with the present invention are mi~
and mouse cell linss heterozygous for the same ~kB deficiency.
Such genetically modifed mice are useful in model systems for 3 ~ DC30 studying human diseases involving neuronal degeneration and neuronal cell loss, as well as in screening for genes, proteins, or other compounds that may prevent or impede neuronal cell death or stimulate neuronal regeneration.

.

El~3URE 1 Targetin~ of the tyrosina protein kinasa domain of tha mouse ~kB gene in D3 ES cells.
~ Schematic diagrarn of the strategy utiiized to target 10 the ~SB locus. Verlical black boxes represent the second (K2) and third (K33 exons of the ~kB tyrosine kinase domain. Thick horizontal lines represent ~kB genomic sequences incorporated into ~he targeUng vector pFRK90. Thin horizontal lines represent additional ~rkB ~nomic sequences. Putative cross-overs bstw63n 15 the endogenous ~EkB locus and the ~argeting pFRK90 DNA are indicated as crQssed stipplsd lines. The PGK-1/~Q c~ss~e is indicated by an open box labeled as ~ The HS~K cassette is indioated by a shaded box labaled as Ntk~. The arrows in-~ide both boxes indicate the direction of transcription. The wavy line 20 repre~entbactarial plasmid sequences. Cleavage sitesfor rastriction endonueleases ~mHI (B), .~ dIII (H), ~I (K), and 2~QT (X) are indicated. The ~Q~I cleavage site used to linearize pFRK90 DNA is indicated in brackets. The region amplified by PGR to iden~fy G418R/GancR ES cell clones car~ying homologous ~ -25 recombination svsnts is indicated by a small open box. The 1.4 kbp 2~ m DNA probe use to identify the diagnostic 4.9 kbp HI and 13 kbp ~ DNA fragments (indicated by arrows) is indicated by a small hatched box.
L~. S~uth~m blot analysis of PCR-posi~ivs 30 G41 BR/GanoR ES cell clones. Genomic DNA e~nrac~ed from repr~sentative PCR~positive ~418~/Ganc~ ES cell clones (lanes 1 ~ :
to 4) or untransfected D3 S cells ~lane 5) were digested with HI or ~I as indicated as submitted ~o Southern blot analysis using the probs indicated in (A). The wild type ~B allele r~

generates 4.1 kbp ~mHI and 11.0 kbp ~nI fragments, whereas the targetcd allele yields the diagnostic 4.9 kbp ~mHI and 13.0 k~p KpnI fragments indicatad in (A).
1 (C). Southern blot analysis of a representative litter 5 derived from crossing ~kBTK (+/-) mice. The wild type allele trkB
producss a 4.1 kbp ~mHI fragment whereas the targeted allele produces a 4.9 kbp ~mHI fragment. Half-fllled symbols represent heterozygous (+/-) genotypes at ~he ~B locus. Fillad symbols indicate animals that died. Sexes of the offspring were not 10 determined and are shown as diamonds.
FIGURE 2. Expression of ~B receptors in ~BTK mutant mice. Haads (A) and brains (B) were isola~ed ~rom eith~r nowborn (P0) ~BTK (~/+), (~/-), (-/-) mice or adult (Ad~ mice and submitted to Western blot analysis for the expression of the (A) gp1 45~5B
tyrosine protein kinase and the (B) non-catalytic gpg5~5B :~
receptors. Samples were incubated with either (a) an unrela~ed mouse monoclonal antibody; (b~ mouse anti-~p145~kB
monoclonal antibody 18-29.2; (c) rabbit preimmune or (d) immune polyclonal antiserum #104 raised against a peptide corresponding to the 13 carboxy-terminal amino acid residues of the non-catalytic gpg5~B protein (Klein et aL, 1 990a). Immunoprecipitates were electrophoresed in 7.5% SDS-PAGE, blotted onto nitrocellulose membranes and incubated with either (A) a rabbit polyclonal antiserum raised against a peptide corresponding to the 14 carboxy-terminal amino acids of the ~ tyrosine protein kinase receptor that also recognizes gp1 45~B (#42) or (B) antisenJm #104. The migration of gplfl~k~ and gpg5~B receptors is indicated by arrows. Mo!ecular weight markers include myosin (200,000), phosphorylase B (97,000~, and albumin (69,000).
FI~URE~. Abnormalitiesinregionsofthenervous systems of ~5BTK (-/-) mice. (A,B~ Trigeminal ganglion; (C,D) Facial motor nucleus. Scale bar is 50 ~m.
3 ~. Coronal saction through the trigeminal ganglion of a wild type (~/~) mouss, 375 ~m from its anterior origin.

3(1~). Coronal section through the tri~eminal ganglion of a BTK (-/-) mouse, 375 llm from its anterior ori~in.
~ (C?. Coronal section through the center of the brainstem FMN of a wild type (~/+) mouse. The FMN consists of a group of 5 large motor neurons coalesced into severat subnuclei tdashed lines), surrounded by an area of reduced cellular density (small arrows in C).
~ D!. Coronal section through the csnter of the brainstern FMN of a ~BTK (-/-) mous~. The F:MN of th~ ~BTK (-/-) animals 10 has a decreased cellular dansi~y of otherwise morphologically indistinguishable motor neurons.
3 (E). Coronal section through the cen~er of a dorsal root ganglion ~DRG) of a wild type (+l l ) mouse.
FJ~I~. Number of neurons in the trigeminal ~anglion 15 and facial motorneurons of I~kBTK (+/~) and ~ ) mice.
. Graphic representation of the number of neurons ~/- SEM in serial sections of the trigeminal ganglion of ~BTK (+I+
and (-/-) mice. T-test comparisons between the number of ;
~anglion cells in the ~d5BTK (-/-) and (+/+) mice reveals a 20 statistically signi~icant loss of ganglion cells (p ~ .001).

4 ~. Graphic represent~tion of neuron number ~/- SEM
in serial sections of th~ facial motor nucleus of ~CkBTK (+/+) and (~
/-) mice. The FMN of the ~BTK (~-) mice is reduced in both cell number (p ~ .002) and total area compared to the csrresponding ~5 structure of its (~/ ~ ) and (+/-) litter mates.
El~;!.IP(~- Abnormalities in reglons of the nervous systsms of ~BTK (-/-) mice. (A,B) Dorsal root ganglion; (C,D) Lumbar motor neurons. Scale bar (A,B) is 100 ,um; (C,D), 50 ,um.
~ (A~. Coronal se~ion through the cQnter of a dorsal root 30 ganglion (DRG) of a wild type (+/~) mouse.
~- C:oronal section through the center of a DR~ of a $ckBTK (-/-) rnousa.
~ . Coronal section through R~x0d layers 8 (Vlll) and 9 (IX) of the lumbar spinal cord of a wild type (~/+) mouse. Motor neurons are seen as large cells with a dark staining cytoplasm and pale nucleus (arrows).
~a;U Coronal section through a similar region o~ the spinal cord of a ~E~TK (-/-) mouse. Althcugh the gross morphology of the spinal cord and differentiation of Rexed layers 8 and 9 do not appear different irom normal litter mates, ~here is a reduced density of motor neurons in th0 ~kBTK (-/-) mouse.
E~BE~ Numb9r of neurons in the DRGs and lumbar spinal cord of ~BTK ( ~/+) and (~-) mice.
~. Graphic representation of gan~lion cell number SEM in the DRGs ~rom spinal cord thoracic levels 12 through lumbar levels 3 (T1 2-L3) of ~BTK (+/+) and (-/-) mice. Ganglion cell numbsrs in the ~5BTK (-/-) animals compared ~o (+/~) mice are significantly ~dueed (p ~ .05) at all spinal oord levels by approximately 30-50%.
~. Graphic representation of the number of motor neurons +/- SEM in Rexed layers 8 and 9 in the ventral spinal cord ~ - -from lumbar level 1 through sacr~l level 1 (L1 -S1 ) of ~kBTK (+/+) and (-/-) mice. Motor n0urons in the ~BTK (-/-) L1-L5 spinal cord are reduced by approxirnately 35% (p ~ .01~ as compared to ~ ~ n the corresponding neurons of (+/+) and (+/-) mice. -~

Detalle~Des~ript~on o7 lthi~ Inv~ntl~n Deginltion of Terms Tha terms used in this specification are defined as follows. Thesa definitions apply to these terms as used throughout this specitication, unless otherwise limited in speci~ic instances.
The terms "deficient" or "deficiency" as usad with respect to a gen3 refers to an allele altered (e.g., by homologous r~combin~tion, resulting in the insertion of foreign sequences) slich that either no product or only an inoperative fragment of the wild type product can be expressed. A "deficient" allele within this definition may also comprise a gene deletion, wherein the gsne has been clleleted ~QI.Q; a gene disruption, whsrein th~ gene is ~.1 3 i ~

interrupted by another gene or nueleic acid sequence; a partial -deletion, wherein one or more nucleic acid sequenc0s or deleted;
a substituticn, wherein one or mor~ bases are replaced by othar bases; and other suoh mutations as wiil be understood by persons 5 having ordinary skill in the art. Such deletions, disruptions, and substitutions may take place in, for example, the codin~ region, a promoter, or an enhancer.
The ~erm "homozygous" as usad with respect to a gene deficiency refers to a genome havin!~ defici~nt alleles at 10 corresponding loci on homologous chromosomes.
The term "heterozygous" as used with respect to a gene deficiency refsrs to a genome having a deficient allele at the corresponding locus on one homologous chromosome.
Process of preparation Mics hetarozygous or hamozygous for the desired gens daficiency are prepared using tha principles o~ homologous recombination. Homologous recombination has long been known in prokaryotic Sp6Ci9S (for a review, see rend~ in Gen~tic~
(1989)). Its use in mice was pioneered by the work o~ Smithies and 20 t::apecchi. See Smithies, O. ~L~ (1985), Schwartzberg, P. L., ~
, (1989), DeChiara, T. M., ~L~L, (1990) . Bas~d on homologous recombination principles, mica homo~ygous or heterozygous for the desired deficient gene may be cons~ructed by the following steps:
(1 ) isolating the wild-type locus (or part of it) ~or the target gene (e.g., using a cl:)NA probe on a genomic library), (2) modifying or disabling the identified locus by genetic engineering techniques (e.g., by gene disruption), (3) preparing suitable host cells ~e.g., male ES c~lls) capable o~ accepting exog~nous DNA, (4) introducing the deficient gene into the prepared host cells (e.g., by electroporation with a replacemsnt Yector), ,, ,~
(5) selec~ing cslls incorporating the deficient gene, preferably by a process involYing homologous recombination, .;:

(6) injacting th~ identifi~cl cells into early-stage mouse smbryos or blastocysts, (7) identifying chimaric animals (usually male) made from the ES cells, (8) crossing the chimeras with tester animals, (9) identifying offspring hetarozygous forthe desired gena deficiency (USUcllly based on coat COlOf), :: ~`
~10) crossing heterozygous mice, and (11 ) identifying offsprin~ homo7ygous for the desired ~ :.
gene deficiency. .::
The wild-type gane and DNA fragments containing all or part of i~s genomic s~uencas may bs id~ntified by screening a ~ :
mouse genomio library (e.g., from 129 Sv mics) with a proba having part of the sequence of the desired gene. Ulpon : ~ identifica~ion, the tar~et wild-lype g~n~ m~y be modified by .
cenventional genetic engineering techniques, making it "d~ficient"
as defined above. One particularly useful method of modifying the targ~t gene i~ disruption by a marker ~ene, which will aid in selec~ion of cells that have successfully undergon0 homologous recombinaffon ~see below). - ~ .
~arb~brly suitable ho~ cells for the process of - :
homologous recombination needed to mutate the d~sired gene -are male embryonic stem cells. These cells can be successfully cultur~d for a large numbsr of ganerations under ~nditions in ~: which they will not di~rentiate, thus releasing their ability to contrib~ne to all lineag~s, which is an absolute r~quirement to generate a mouse carryirlg th0 mutations introduced in these c~lls.
Ths deficient gene is then introduced into tha host cells in a suitabb manner, such as ~lecliroporation. Onc~ taken up by the host c~lls, homologous ~combination with the corresponding endogenou~; gene may occur.

~ 0-To select cells in which the desired homologous recombination event has taken place, a marksr may be used. -Suitable markers include genes conferring resistance to such antibiotics as neomycin. For example, a bacterial nQQ gene 5 confers resistance to neomycin and such analogues as G418. The :
marker gene may be inserted in thls target wild-type gene, thereby disabling the target gene while providing a selactable marker for cells having taken up. Among the cells that survive treatrnent with neomycin or one of its analogues (i. a., cells that are ~I~Q+), a 10 limited fraction will have replaced thc wild-type gene with the desired deficient gena.
The deficient gene may, however, enter the host genome : -by random insartion instead of hamolo~ous reoombination. Such nonhomologous r0combinants will be ~ without the dasir0d 15 repiacernent of the wild-type gene. Therefore, to fur~her salect homologous recombinants, the ends of ths deficient gene may have such negative markars as the herpes simplex vin~s thymidine kinase (I ISVTK) gene. The HSVTK gene product converts gancyclovir into a ~oxic metabolite. In a homologous 20 recombination event, however, th~ HSVTK gene will not be present. Thereforo1 desired rscombinants will be resistant to gancyclovir and may be further sel0cted by g2ncydovir treatmQn~.
After biological resolution, still most recombinational events will be of the non-homologous type (in some cases, as 25 many as 99% will be non-homologous ~ecombinational events).
To identify th~ tnJe homologous recombinan~s, molecuiar s¢reening is nacessary either by Southern blot or PCR. This technology has been pr~viously described (Joyner t al., 1989).
When a suitable ES cell clone has been identified, the 30 host cells may be injected into early-stage embryos or blastocysts, and reintroduced into a pseudopragnan~ fsmale. Chimeric animals will generally result from at Isast some of thsse embryos, their tissues deriving in part from th~ selected clone. In some of these chimeras, the ES c~ll clone carrying the targeted gene will contribute to the germ line. If so, the naxt generation of animals ~;
will bs entirely derived from the ES cell clone selected in~i~. af~er -homologous recombination as described abovs.
The heterozygous progeny can be cross-bred to yield homozygous animals. Confirmation o~ the allelic structure of the . : - i ~- -mice oan be ascsrtained, for exampl~ by PCR and Southern .~ ~ :
blotting. -:
:' Using recombinant DNA t~chnology, a targeting vector carrying an aKered ~kB genomic sequence was constructed. This vector was used to destroy the endogsnous ~sB loous of ES cells by a process involving electroporation of the v~ctor DNA ~ollowed by homologous r~combination with endogenous i~kB sequences.
The pref~rred mutation elimina~es expr~ssion of gpl 45~B, the signaling r~ceptor for BDNF and NT-4, without c eliminating expression o~ the non~atalyUc isoforrn, gp95~kB Mice homezygous for this mutation, ~kBTK (~-3~ develop to birth.
Howsver, these animals do not display any signs of ~eeding ac~ivi1y and most dis at P1. Neuroanatomical examination of these mice reveal~d significant n~uronal defic~encies in the central lfaaal motor nucleus and spinal cord) and peripheral (trigeminal and dorsal root ~anglia) nervous systsms. However, qualitative examin~tion of ather neural structures (cercbral co~t~x, hippocampus) known to express ~kBTK transorip~s do not appear to be affected, perhaps due to compensatory mechanisms. These findings illustrate the critical rob of the gpl45~k~ tyrosine pro~ein kinas~ r~c~ptor in th~ mammalian nervous systsm Tar~tln~ th~ Mouse ~kB Locus: E~xperlmental Appro~h A mouse genornic library derived ~rom NIH3T3 eells was soreened with a 2.7 kbp ~II ~B cDNA probe eneompassing sequenc~s encoding the transmembrane and cytoplasmic domains of ~pl 45~kB. One of the library phage was ~ound to : .
contain a 21 kbp in~ert which inciuded the second and third ~xons I~C30 of the ~yrosine protein kinase region of ~[kB separated by a 6 kbp long intron (Figure 1). Since the exon/intron structure of the ~CkB .
locus has not been fully established, we will r~fer to these exens as K2 and K3. Exon K2 is 173 bp long (nucleotides 2225 to 2397 5 of pFRK43; sse Klein ~" 1989) and contains domains lll through V of the tyrosine protein kinase region (Hanks QLaL
1988). Exon K3 is 235 bp long (nucleotides 2398 to 2632 in pFRK43) and contains domains Vl and Vll.
ThQse g~nomic se~uences were used to constn~ct the 10 replacemQnt vector, pFRK90, by replacing 33 bp of axon K2 (nucleotides 2330 to 2362 of pFRK43) with a PGK~ Q cassette (McBurney ~L 1991 ) inserted in the same transcriptional orientation as the ~kB gene (Figure 1). The shor~ arm of pFRK90 was genarated by PCR-aided amplification of phage genomic 15 sequences and consis~s of 104 bp of exon K2 and 750 bp of upstream intronic ssquenees. The 7,250 bp long ann of pFRK90 was located 3' to the PGK/D~Q cassstte and contains the remaining sequences of exon K2, the 6 kbp long K2/K3 intron, sxon K3 and 1 kbp of downs~r~am intronic sequences (Figure 1).
20 A thymidine kinase cassette, used for negative selac~ion o~ cells carrying non-homologous recombination (Mansour~L 1988), was insert~d 3' of tha genomic ~zkB sequsnces. :
pFRK90 DNA was lineafizad with ~Q~ and eiectroporated into 5 x 107 D3 ES cells (D3 clone) as descnbed in :
25 Exarnple 1 hereinafter. Two days later, the elactroporated cells wsre placed under dual salection in the presence of G418 and gan~clovir. This doubb sel~ction protocol resulted in a 10-fold enrichment over G418 sebction alone. A total of 800 G418R/GancR deubl0 resistant D3 ES cell clones ware picked and screened by PCR as descnbed (Joyner ~L~L 1989). Nine ES
cell clones found to be positiv~ in the initial PCR scrsen were subsequently submittad to Southem b~ot analysis. As illustrated in Figure 1 B, a probe dcrived from intronic sequences locat~d ups~rsam of pFRK90 detect~d a 4.9 kbp ~m~I (4.1 kbp in wild type trkB DNA) and a 13 kbp ~Sen~I (11 kbp in wild type trlcB DNA) DNA fragmen~ diagnostic of homologous recombinational events in eight out of nine PCR positive ES cell clones. These results indicate that the targeting frequency was one clon~ in 90 G418R/Ganc.R double-resistant ES cells.
C;eneratiorl of trkl3 Mutant Mice Targeted ES cell clones wera injected in~o C57BU6J
blastocys~s and Iransfarred into the uteri of psaudopregnant CD1 recipient mothers (see Example 1 hareinafter). Of a total of five CIOneS injected, three of them (K2- 1 9A, K2-24 and K2-29) generated chimeric offspring with E S cell contnbutions ranging from 20 to 90 % as judgsd by the proportion of agolni coat color.
Three chimeric males d~riv~d from the K2-29 clone exhibited greater than 50% agouti ooat color. Thase chimeras were bred to C57B116J mice (as well as to 129Sv mice) and found to transmit the targeted allele through the germ line. Breeding of two chimeric siblings derived from an independent ES ceJI clone (K2-19A) which displayed w~ak to modera~e (20% to 40%) proportion of agouti coa~ color also resulted in gerrn line ~ransmission o~ the targeted ~kB gene. Unlass otharwis0 s~ated, the results described in thi~ study wer~ obtained with mice derived from the K2-29 ES
call clone.
Genotypin~ ot the agouti offspring producad tho sxpected ;
frequency of 50% h~erozygotas, These mice, trom now on -~
designated as ~BTK (~ grew normaliy and showed no -~
obYious anatomioal or behavioral dsfects. To study the phsnotypic c~nsequences of elimination of a func~ional ~S~ ~yrosine protein kinase, we crossed ~hese ~kBTK (~l-) heter~ygous animals. The rasulting lit~ers were nonnal in size and all the animals appear3d : :
norrnal at tha time of birth. G~notypic analysis of tail biopsies from 116 offspring mice showed that homozygous trkBTK (~-) animals were bom at a frequency o~ 23.3 % indicating that mice lacking gp145~B r~ceptors can d~velopto birth (Table 1). Figure 1C
depicts a South~rn blot analysis of a represantative litter.

fl~

TABLE 1 Genotypic analysis of ths offsprin~ of ~kBTK ( ~

GENOTYPE NUMBER PERCENTAGE
~ QF MICE _. . - .
(+ / +) 33 2~.4 %
(+I-) 56 48.3%
~-L-~ 27 23.3 %
Expresslon of trk~ Receptors To verify that tha targeted L~SBTK ~_/O) animals did not express the catal~ic gp145~kB receptor, we immunoprecipitated protein Iysates obtained ~rom heads of newborn wild-type (~I+), ~,d5BTK ~ 1/ 3 and ~kBTK (~-) animals with a monoclonal antibody elicited 2gainst the tyrosine protein kinase domain of gp~45~ .
The resulting immunoprecipitates ware submitted to Westsrn blot .
analysis using pan anti-~k antibodies. As shown in Figure 2A, the homozygous ~BTK (j-3 animals did not show da~ec~able gp1 45~kB protein whereas the ~BTK ( I /-) heterozygous displayed reduced lavels of this raceptor. Similar r~sults were obtained in parallsl experim~nts in which the protain Iysates were immunoprecipitated with another ~kB specific antiserum raised against a pep~ide corr~sponding ~o amino acids 794 808 of the -~B raceptor (Klein ~L 1989). ~: :
Tar~ting of ~B genomic sequences encodin~ ~he tyrosine protein kinase domain of gp14~kB should not disrupt ~: expression of the non~atalytic gp95ltka r~oeptor (Klein ~L
1 990a). To confirrn that the phenotypic properties (see beiow~ of the targeted i~kBTK (~_) mics wer~ du~ exclusiv01y to the elirnination of the gp145g~ tyrosine protein kinass receptors we tested for the pr~ance of gpgBlk~ in these animals. As shown in Figur~ 2B, the homozygous ~BTK (~-) mic~ retained the ability to express ~he non~atalytic gpg5~kB receptors at levels comparable to those of ttleir hetero2ygous littQr ma~es.

Phenotyp~c Analysis ot ~rkBTK (-l-) Mlce As indioated above, all offspring derived from crosses of heterozygous ~sBTK (+/-) mice appeared normal within a few hours after birth. The first symptomatic differ~nce could b0 observed at approxirnately lwelve hours when some of ~he newborn animals wera found to be without milk in their stomachs.
Most of thes~ animals died by P1, although some suNived as long as P3. Subsequent Southern blot analysis of tail biopsies established tha~ these mice were homozygous for th~ ~IsBTK
1 9 mutation.
Gross examination of these ~kBTK (-/-) animals at the day of birth revaaled that they were the same size ~crown-mmp) as thcir unaffec~ed lit!er m~es and did not exhibit any appar~nt physical deforrnities. Sinoe these ~sBTK (~-) animals did not taka nourishment, they wer~ chscksd for abnorrnalities in thair ~ -digestive sys~am. No gross lesions were obssrvad in the head, . ~:
including clefl lip or pala~e, although in a few instances thsre appeared to be a slight macroglossia. The ~sophagus appeared ~ :
normal and without stricture, as did the s~ornaoh an~ pylorus. One differen~ obsen~ed in tha ll kBTK (~-) mice cornparad to their : : :
~+1+) and ( ~ ter mat~s was the oc~sional occurrenca of a gas-expanded s~omach, most likely due to an absenc~ of matemal -milk.
A rudimentaly neurological exam was performed on :
th~se n8WbOrll mia~ to try to identify gross behavioral differences between the ~E~ 3~ ) and (~-~ animal~. W~ found that upon a light st~olcin~ under the chin, the (+t~) and (+/-) animals :
r~spondsd by orienting to the side of the stimulus. These animals also responded by opening and closing their mouths in what might be interpreted to be a sucking pattern. This behavioral paffem was not obsen~ed in the ~SBTK (~-) mice. These animals did no~ orient towards the stimulus and for ths mos~ part never open~d their ~ -mouth except for an occasional gasp. Subjsctively, it appeared that th~ ~BT~ ) and (+/-) animals had a greater number of 4 ~ ~ DC30 vocalizations. All other parameters of behavior appear~d normal.
For example~ all of the mics ~xamined wriggl~3d in response to being handled. Homozygous ~IsB~I~ mice derived from the K2-19A
ES cell line also exhibited the same phenotype.
It was not clear whether the lack of milk in the stomachs of the ~IsBTK (~-~ mice was due to the mothsr refusing to fee~ . -them, or to an intrinsic defsct in these animals. To address this qu~stion, manual feeding the trkBTK (-l-) nawborns was ~ :~
attempted wi~h milk formula through a syringe attached to a small calib~r tube inserled into their mouths. Whiie this procedure was su~essful with the (+l l ) and (~/-) mic~, the ~[kBTK (-l-) animals inhabd instead ot swallowed the milk.
The ~BTK (-/~) mice did not develop properly after birth, showing clear si~ns of cachexia and retarded development by P2, presumably due to their inability to properly feed thems21ves. As :
indicated above, most of the ~kBTK (~-) mice died ~ P1. Those that occasionaliy sun/ived were found to be sev~ly cachectic and died of marasmus. To avoid the possibili~y that the obsarved anatomical abnormaiities might have been due to wasting, animals wer~ ~xclusiv01y analyzed at P0.
lTK (-/-) M5ce 11ave Leslons in the Nellronal Systsms Involv0d in F13&âi Feeding behavior in mammals is controlled hy complex int~racffons arr ong several neuronal pathways. The sensory system, for the most part, is subseN~d by lhe maxillary and mandibuiar branches ot ths tng~minal nerve. The motor system is controllQd by branohes of the ~acial nen~e with contributions from the mandibular branoh of the trigeminal ne~e (Walton, 1977). To det~rmine wheth~r there was any physical defi~it which would underlie the observed behavioral abnormalities, we axamined tho sensory ganglion of the trigeminal n0rve and the mo~or nuclsus of the facial nerve of these ~kBTK (~-) mic0. Both of thase structures have been previously shown ~o axp7~ss ~kB ~ranscripts during developm01lt ~KIein ~L~L 1989). Upon gross exarnination, tha lrig~minal ganglion of the ~ BTK (-/-) mice appeared smaller than ~hos~ of the nonTlal lilter mates (Figure 3A, B). Microscopic axamination revealed a significant (p c 0.001~ reduction in the -number of ganglion cells present in the ~sBTK (-1_) mica (8,469 +
5 698; n~4) compared to the (~/~) (21,132 + 567; n=4) and (~
(20,314 i 927; n=3) animals. The number of ganglion cells found ~ ~ :
in the (+/+) animals comparas well with that reported in a previous study (Daviss and Lumsden, 1984). The largest difference in ganglion cell number between the ~BTK (-/-) and (~/+) mica was observed in the antenor on~-half of the ~anglion (Figure 4A).
Howevcr, there was no detectable difference in the size of the ganglion cells of the ~dsBT~ ) mice whcn compared ~o the (~/+) or (~/-) animals.
The nucleus of the Sacial nan~e lies as a discrete group of :: :
cells in the rostroventral brainstem, which enenrates ths ~ .
musculatur~ of the head and neck. This nucleus is characterizsd ~ ~ :
by a dense ~roup of large neurons surrounded by a halo of reduced cellular dens~y (Figure 3C, D). Qualitative comparisons between the I~kBTK (~-) mice with the corresponding (+/ ~) and ~ .
(+/ 3 animals show0d that the wild-~ype faaal nucl~us had a greater density than that of th~ ) mu~ant mics (hgure 4B).
Subsequent counts of facial motor neurons in the ~kBT~ I+) and (~-) mic~ showed a significant difference between the two .
gr~ps. Wh0r~as the ~kBTK ~ ) mi~ had 3,291 ~ 357 ncurons (n~4), the (~-) mice only had 1,019 + 65 (n34) (p ~ 0.002~. Thase ~llular dafi~i~ncies encompass each of the nuclei's subdivisions.
The number of motor neurons in the (~i ~) mice correspond to those repcrtad by Herrup ~" (1984). ~:
Other N~uronal l:laflclen~ies Ir~ ~kBT~ ) mlce -:
BDNF, a oognate ligand for the ~B eceptors, oan support th~ gr~wlh and survival of neurons from DRG v~ro (Lindsay ~, 1985; Kalcheim ~L 19~7). To examine whether the ~.[kBTK (~-3 mice had any addiUonal neuronal defects in thas0 struc~ures as a consequenc 3 of loss of gpl45~kB expression, we counted DRG cells from the T11 to the L3 region of the spinal cord. ~r As illustrated in Figure 5A, we observed an approximately 50%
decrease in the number of calls in the DRG. Th3 DRG cells counted from the ~BTK (+/-) animals segregated with ~h0 (~
5 wild type mice. The loss in cell number in these gangJia is also apparent by their significantly redLIced size in the ~k8TK (-/-) mice (hgure 3E, F).
Recent studies have indicated tha~ BDNF can prevent cell death of axotomixed motor neurons in newbom rats (Sendtner ~
10 ~ .1992; Yan~L~L, 1992). Moraover, BDNF has also been shown to prevent naturally occurring and differentiation-induced cell death of lumbosacral motor neurons in chick embryos (Oppenhaim ~L 1992~. Thsrefore, we examined the number of lumbar motor neurons in the spinal cord of IlkBTK (+/+) and (-/-) 15 mice (Figure 3 G,H). Tha number of motor naurons from the (+/+) animals were significantly (p < 0.01 ) higher (2,552 i 97; n=4) than tha~ counted in ~he ~kBTK (~-) mice (1,667 ~ 72; n=4), parlicularly in thosc neurons located in spinal cord levels L2 to L5 (Figure 5E~).
No differQnces were observed in the sacral S1 r~gion (Figure 5B).
20 The number of mo~or n~urons prssent in the IdsBTK (+/-~ mica wera not statis~ically different from those counted in the (~/+) animals. Despite ths cell loss seen in the ~kBTK (-/-) animals, the rsmaining lumbar motor neurons appear identical to thos~ of their ~nld typ~ litt~r mates (Figure 3G, H).
i~kB transcripts have b9~n described in a vanaty of other -n~ural structures includin~ the cerebral cortax, the pyramidal csll layer of the hippocarnpus, thalamus and Purkinje ~lls (Klein ~L
1989; 1990a, b). Quali1~tive microscopic Qxamination of thes2 r~gions did not reveal any detectable changes in ~he ~kBTK (~_3 mice whsn compared to ~heir norrnal (+/~) or (+/-) litter matss.
Dlscusslon 11-he preferr~d embodiment concerns transgenic mice lackin~ a ~nctional gpl45~kB tyrosine protain kinase receptor by specifically targeting those L*B sequences encoding its catalytic i 4 ~ ~ DC30 ~ ~ ~
,~ 19 kinase domain. This defact is likely to disrupt trophic signaling by its twe primary ligands BDNF (Klsin ~L. 1991 b; Soppet ~, 1991; Squinto ~ al~, 1991 ) and NT 4 (Be~<emeier QLiaL., 1~91, Klein ~L, 1992; Ip ~, 1992). 9p1 45~5B call also serve as a 5 receptor for the related NT-3 (Glass ~ , 1991; Klein ~L~, 1991b) but only when ectopically expressed in non-neuronal cells ~Ip ~L, 1993). Therefore, it is unlikely that disruption of gp1 45~B
expression in the ~sBTK (~-) mi~ has a slgnificant effect on NT-3 signaling in vivo The ICkB locus encodes a second class of neurotrophin ~:
re~ptors iacking the kinase catalytic domain (Klein ~L 1990a;
Middlemas ~ 91). At least on~ ofthese receptors, gpg5i is expr~ss~d at high levels in the adu~ mouse brain, particularly in ~he ependymal layer of certain cranial vantricles and in the choroid 15 plexus (Klein ~L. 1 990a). Our targeting stratagy was aimed at avoiding disnJpUon of this receptor. Indeed, Wes;tem blot analysis of brain 9xtrac~s obgain3d from P0 ~BTK (-J-) mice r~vealed that whereas ypl 45~3 is undet~able, gpg5~B is expressed at levels somparable to 2hose of the (~/~) and ~+/-) mice. Tharefore 20 the phenotypic abnonnalities obs~n~od in the targeted ~kBTK
mice are lik~ly to b~ due to thn sp~cific disruption of signaling throu~h ~he gp145~B tyrosine protein kinase raceptor. : ~-At birth, the ~kBTK (~-) mice appear morphologically indistingui~hable trom their (~1+) and (~/-) siblings. t lowever, the ~ -25 (-J-) mics do nDt show any signs of feedin9 activity as datermined by the absen~ of milk in their stoma~hs. ~U P1, ~he IlisBTK (~_) mis0 are already smaller in size and many die. A~ P2, occasional survivors depic~ clBar signs of cachexia. So far, none of the ~kBTK
(~-) mice has sunAv0d beyond P3. The inability of thase ~argeted ~:-30 mic0 to fead is likely to be a consequQnce, at least in part, of tha signi~lcant rsduction o~ cells in the trigeminal and facial nuclear systems (Chusid, 1~73). Th0s~ neuronaJ de~ciendes may also account for the observed inability of the ~BTK ~_) mice to respond to simpb esnernal stimuli such as touching their faces or - ~ - 20 -~entle stroking under their chin. These abnormalities were observed in mice derived from two independent ES cell clones, further indicating that ~argeting of the ~B locus is direc~ly responsible for these deficiencies.
The trigeminal ganglion of the trkBTK (-1-) mice only contains 40% o~ nsurons present in their l~/+) or (+/-) siblings. In ~iILl. hybridization analysis of mouse embryos has shown an intense and rather homogenaous expr0ssion of ~kB transcripts in this ganglion at E9.5 (see Figur~ 2 in Klein ~L, 1990b).
However, at E14.5, ~sB expressioll appears spot~y and in less than half of the cells (see Figu~e 6 in Klein ~L, 1989). Neuronal cell loss in the trigeminal ganglion does not appear to be evenly dis;tributed, wdh most of the missing cells corresponding to thos~
derived from ths anterior portions of ths gangiion. A possible explanation for ~his observation comes Srom embryological studies in chickans, which indioate that the anterior par~s of the ganglion are derived frorn the non-NGF~ep6ndant ectodermal placodes (D'Amico-Martel and Noden, 1983). Al~hough tho duality of trigeminal origin has not b~en conclusively established in marnmals, th~ pattern o~ cell loss seen in ~he homozygous $~kBTK
animals appcars to conform to the distribution of fibers and cells th~t conlnbute primarily to the mandibular and ma~cillary branches of the mammalian trigsminal nerve (Erzun~mlu and Ki31ackey, 1983). In this s~udy, we only examined the peripheral control of senso~y fun~ion in the faaal region of the ~BTK (~_) mio~. We do anticipa~e however, that CNS regions ~hat subserve ths ~ame s6~nsomotor functions may also be affscted. This hypoth~sis is supported by a recent study indicating BûNF
d~pendence of mesencaphalic trigeminal nourons (Von Bartheld and E~o~hwell, 1993).
The effe~ent limb for feeding behavior is controlled by motor neurons locatecl in the facial nucleus as well as by ths mot~r flbars of the maxill~fy division of the tngeminal n~rva (VVatton, 1977). The significant loss of motor neurons in ~he faaal nucleus (up to 70/c~) observed in the ~IsBTK (~-) mice is likely to disable their mastication muscles and tharefore ~use th~ir apparsnt inability to suckle. Theso findings are in agreement with the r~cent observations of Senc tner ~L~ (1992) indicating that BDNF, one of the cognate ligands of ~ap~45ILkB, can pravent dea~h of facial motor neurons after axotomization of the facial nerve in newborn rats.
The protectiva activity of BDNF on faciat n~urons following axotomy is not restrictsd to those of the facial nucleus. Recent studies have indicated that this neurotrophin also has survival promoting effects on spinal motor naurons following transection of the sciatic nerve in newborn rats (Yan ~, 1992) and can resaJe :
chiok embryonic motor neurons from naturally occurring cell death (Oppenhsim ~, 1992). The significancs of th~se observations is underscored by our resu~s ~i~th the ~kBTK (~-) mice, which demonstrate ~hat signaling though the gpl~B receptor, either by BDNF or by NT-4, is an absolute r~quirament for survival of at least 1/3 ot the lumbar spinal motor naurons.
kL~. hybridization studies have indica~d that ~sB is ~undantly expressed in the spinal csrd throughout embryonic d~valopment (Klein ~, 1989;1990b). Yet, gross examination of ~ -spinal oord cells does not reveal dramatic differences between the ~kBTK (-J-) and ( ~ miee, other than in the mo~or neuron population. A similar observatien has been made in other parts of the mouse nervolJs sys~em known to exprass high levels of ~E~
transcripts such as the cerebral cortex, the pyramidal csll layer of ~he hippoc:arnpus and the thalarnus. It is ~ssible that some defidencies Ylnll be tolJnd in these structur~s after more detailed analysis. Howevar, it is also possible that oe~ain gp14~-expressing n~urons may sunrive in the absence of this receptor thanks to compensatory meehanisms, p0rllaps provided by the highly related ~kC tyrosine protein kinasa r~cep~ors. Indesd, Id~C
tran~eripts are also abundant in the spinal cord, cer~bral cor~ex and hippocampus (Lambalb 5~, 1991 and submitted; Msrlio ~1,. 1992)- The recent availability of mice carrying a targeted l~kC

~30 gsne should help to establish the relative contributions of these tyrosine protein k~nases te neuronal survival in thasa structures.
Interestingly, one of the stnJctures showing obvious deficiencies in the ~jBTK (-/-) mice are the DRGs, which are 5 known to express transcripts from each of the three known rnembers of the ~k gene family, ~s, hkB and ~kC (Martin-Zanca ~L 1990; Klein ~L. 1 990b; Ernfors ~1~. 1992; Lamballe submit~ed). Ye~ in this study, about 50% of their neurons are absent and their overall size is considerably smallar. A possible 10 sxplanation for the observed c~ll loe;s is that ~ach of the m~mbers of ths ~k gene family has a distinct function and do not ~mplement each oth~r. Alt~rna:tivaly, each of these genes might be individually expressed in speciflc sub~ots of neurons rendering them responsive to speafic memb~rs of the NGF neuro~rephin 15 family. In support of this hypothesis, DiStefano ~aL. (1992) have observed that radiolabol0d N~;F, BDNF and NT-3 recognize differ~nt neuronal subpopulations in adult DRGs. Moreover, cultivation of E1~ rat DRG neurons in th~ pre~nca of either NGF, BDNF or NT-3 results in the s~rvival of c~lls specifically 20 exprQssing ~, ~ orI~C mRNAs, respeclivaly (Ip Q~ aL, 1993).
Thes~ observations suggest that only a vary limit~d num~r of DRG neurons, if any, sxpress more than one member of ths ~ : -gene family of r~ceptors. If so, the observed cell loss in these gan~lia is Ukely to corr~spond to that subset Qf neurons that only 25 ~xpr~s~ ~aP~45~-Re~a~less of the signihcant abnonnalities obsQrv~d in :;
the DRGs ot i~BTK (~-) rnic~, it is unlikely that they eontribut0 to ~ ~-their dsmiss, sincæ none of these mice have surviv3d beyond P3.
Additional ~tudies will be necessaly to evaluate tl~ full extent of 30 neuronal defiaencies caLlseb by the disruption of gpl45 expression, sinc0 many det0cts may not display obvious pheno~ypic d~fioi0ncie~ in such young animals. Of particular interest will b0 the analysis of the developing substantia nigra, a stmcture in which ~he prot~ctive eff0~s of BDNF to chemical .
"~'"' ;''''~' . . ..

3 ~

~~ - 23 -insults have been already illus~rated (Hyman ~L~L, 1991; A~ar .L, 1992~.
The neuronal cell loss observed in the ~BTK (~
animals might be due either to abnorrnal deveiopmental diffe~entiation or inadequate survival. One method to discriminate between thsse two possibilities is to examine each ot the struc~ures known to have neuronal def0cts in the ~ckBTK (~_) animals after neurogenesis, durinçl ths period of axon ingrowth or naturally occurring cell death. Two regions whare we observed cell loss in the ~BTK (-1-) animals have been analyzed at this critical period of davelopment. They include the trigeminal ganglion at E12 and the motor nsurons of the spinal cord at PO. In each case, we found many more pyknotio and fragm0nted nuclei in the ~kBTK (~ mic0 than in their (~/~) or (+/-) siblings. Thase observations suggest that a major component of the neuronal cell loss ssen in the ~kBTK (-/-) mice is due to incraassd cell death.
These findings are in agreement with the studies of Vogel and DaYiss (1991) indicating that the onsat of BDNF dependence might be coordinated with target en0rva~ion. It is not known at this tima H additional defects during neuronal differsntiation may also contribute to the obseNed phenotype in thes0 ~kBTK ~_) mutant mice.
It has been proposed that trophic signaling through the ~p145~ Wnase r~ceptors roquir~s the pr0sence of the low affini~ neurotrophin receptor, p75LNGFR (Bothw~ll, 1991). Support for this h~pothesis came from the studies of Hempstead ~3L.
(1991 ) who report~d that the r~lat~d I~ r~ceptors required co-expression of p75LNGFR in order ~o gensrate high affini~y NGF
binding sites. The resul~s of Soppet ~L (1991) indicating that -~
~O gp145~ re~ptors alone also bound BDNF with low affinity in the : :
nanamolar range, providad furth~r support for this hypothesis.
Oth0r studi~s how~ver, have shown that gp1 45~B receptors could mediate BDNF and NT-4 signaling in the absence of p75LNGFR, albeit in non-n~uronal ceils (Glass et aL, 1991; Ip ~L 1992, - . :

i 4 ~ ~ DC30 1993; Klein et al~. 1991 b, 1992; Squinto ~L, 1991). More racently, Marsh ~" (1993) have illustrated ~hat gp1451rkB can also signal in cultur~s of hippocampal neurons which do not ~xpress p75LNGFR ~ceptors.
Now, genstic studiss strongly argue agains~ the gp145~SB/p75LN~iFR heterodimer r~3csptor model (Bothwell, 1991).
Unlike th~ ~BTK (~-) animals d~scribed in this study, homozygous (~-) mice carryin~ a targeted p75LNGFR gene deveiop normally and only dispiay obvious neuronal defsc~s in the t 0 sensory enervations of th~ footpad skin (Les ~L, 1992). Indead, since p75LNGFR serves as a reoeptor for each of the four known mambers of ths NGF neurotrophin family (Rodriguaz-Tebar ~, 1990, 1992; Hallbo~k ~" 1991), the gp1 45~kalp75LNGFR
heterodimer model pr~dicts ~hat the absenoe of p75LNGFR would ~ .
r6sult in a more sevare phenotype than muta~ions in any of ~he individual members of ~he ~k gene farnily of kinase receptors. Our ~ndings however, do not rule out the possibi!ity Shat p75~.NGFR may play an auxiliary rob in gp145~ signaling. Crossing th0 ~kBTK
) and p75LN~;FR ~ ) mice should provide valuable information regarding the contributions o~ these two classes o~ :
neurotrophin rac~ptors to the d~vslopment and maint~nance of ::~:
the mamrnalian nervoussystem.
The invention wiil roow be fur~her de~cribed by the followin~ w~rkin~ examples, which ars preferr~d embodimen~s of the invention. These examples are illustrative rather than limi~ing.
Unless other~nse indicated, all tempsra~u~s are in degr~es Gelsius (C). Although the following specific examples all concem ~:kE3. those having ordina~y skill in the art would be able to adapt ;
these procedures to other members of the ~ iamily of re~ptor~
~.'' ;:

.. , . . , . .,. ,. ~ .. . .. . . ....

4 ~ ~ DC30 - 2~ -Ex~np!e 1 Tar~etlng vec~or Tha targeting vec~or, pFRK90, consisted of 8.1 kbp of ~kB
genomic sequences (850 bp in the short arm and 7.25 kbp in the 5 long arm), a phosphoglycerats kinase-1 (PGK-1)/~ cassette inssrted wi~hin ~kB coding s6quene~s and a flanking HSV
thymidine kinase (tk~ cassette (FI~ure 1). To ganerata pFRK90, we first screened an NIH3T3 mouse genomic library with a 2.7 kbp ; ~:
~II cDNA fragment of pFRK43 (Klein ~L, 1989) ancompassing those sequences encoding the transmembrane and cytopIasmic domains of gpl45~kB. One of ~he n3combinant phages (#12) contained a 21 kbp long insert which included the second (K2) and third lK3) exons of tha ~SB tyrosins protein kinase region. A
4.8 kbp ~dm DNA fragme~ was used to generate the short arm of pFRK90 by PCR-aided amplification using as amplimers a 5' primer (SEO. ID. NO.: 1 ) having the -~equence 5'-CCTT~j9~TCTTCAGAAmATTMAGAG-3' which annealed to intron sequences 750 bp upstream of exon K2, .
and a 3' primer (SEQ. ID. NO.: 23 20~ 5'-GTCGCC~Çg~ACAGACACCGTAaMCTTG-3') --~
tha~ anncaled ~o ~xon IC2 sequences (nucleotides 2311 to 2341 o~
pFRK43). The ur~erlin~d s~quenc~s oorrespond to ~ (5' primer) and ~hQ;l (3' primer) cleavage sites used for subelonin~
purposes. This 850 bp ~ hQI PCR-amplifled DNA tragment ~ -was subcIoned into pBlu~script along with a 1.9 kbp 2~
~GK-1/D3aQ casset~a d~rivad from pKJ-1 ~McBurn~y ~L 1991).
The resulting 2.75 kbp ~ ~I DNA fragment was subs~quently ligated ~o a 3.85 kbp ~ ~,l DNA fra~ment containing ~he 3' 35 bp of exon K2 followed by 3.8 kbp of downstream intronic sequences. The ~ cleava~e si~e of this 3.85 kbp ~-5~ DNA
fra~ment was engineered by PCR-aided amplification of sxon K2 sequences in a manner that aliminated 33 bp (nucleotid2s 2330 to 236 of pFRK43) from sxon K2. Ths 6.6 kbp f~ ,I ~B DNA
fragment a~ntaining the PGK-1/neo cassatte inse~ed within exon - .

K2 sequ~nces was next subcloned into the ~ I sites of pFRK75, a pGEM-9Zf(-)-derived vector containing the HSV tk casset~e of pMC1TKpA (MansourQ~I" 1988). The r~sulting plasmid, pLL41, was utilized to gen~rat~ th~ tar~eting vector 5 pFRK90 by addin~ an additional 3.4 kbp ~lal genomic DNA
fragment Qf phage #12, corresponding to thosa sequances immediately downstream from the 3.85 kbp ~-~:[ DNA
fragment, ~hus inoreasing to 7.25 kbp the total length of the long arm of pFRK90.
10 Transfectlon of ES cells and E~lastocyst Inlection Cell culture and electroporation of male D3 ES cells ~:
(Doetschman ~, 1987) were essentially done as describad (Wurst and Jayner, in pr~ss). ES cells wer0 trypsiniz~d, washed in PBS and electroporated with 40 ,u~ of ~I-Iinaarized pFRK90 p~r 5x1 o6 cells USiR9 a Bio-Rad Gene Pulser (240V, S00 ~ . Cells : ~
were seeded onto 100-mm gela~inized culture dishes at a density ~ ~ -of 2.5x106 cells par plate in ES cell culture m0dium containing :
15% f~tal calf s~rum and 500 UtmL of LIF. Aftar 48 hours, celis ~:
were subjected to double sel~ction with 250 ~g/mL of G418 and 20 2.2 ~1~ gancyclovir. Colonies were picked 10 days after transfection usin~ the half colony m~thod as dsscribed (Joyner i~l" 1989). Posi~ive o011 clonas (see bebw) were picked and transferred onto a monolayer of mi~omycin C-tr~ated rnouse embsyonic feedsr c011~ in ES cell medium without G418 or LIF. For 25 blastocyst inje~ions, calls were trypsinized, washed in PBS, and kept on ice. Approximately 1~ cells were injsctgd into C:57BI/6 blastooys~s as described (Joynar ~t al~. 1989) which were then transf0rred into th~ utenls of pseudopr~gnant CD1 females. The resulling chimeras wer0 bred onto a C57BU6J backgr~und.
30 PC:R Screer~ nd Southerrl biot an~ly31 Pools oS 20 individual G41 8R/GancR ES ~ransformants ~ .
wer~ tested f~r homologous recombination with pFRK90 DNA as ::
describ~d (Joyner ~ 1989). Briefly, approximately 104 cells - ~ -wer0 Iysad by frs~zin~ and thawing in deionized water and tr~ated : -~13 ~ DC30 with Proteinase K for 2 hours at 50C. Half the sample was submitted to PCR ampliflcation L94C: (1 minute), 65c (2 minutes and 72C ~3 minut0s) for 40 cyclesl in the presene~ of 1.25 m~
MgCI~ and 10 m~ NTPs. The 5' amplimer (SEQ. ID. NC).: 3), 5 having the sequance 5'-GCTGGACACTGGGACTGCCAGGCC-3' corresponded to genomic sequena3s located 20 bp upstream of ~he 5' and of tha short arm of pFRK90. The 3' amplimer (SEQ. ID. ~ ~ ~
NO.: 4) having the sequence - -1 0 5'-CTACCCGGT~Ç~ge~g-3' contain~d ths s~QRI and 2~hQI cl~avage sitss (und~rlined) located : i :
at the junction be~ween the exon K~2 ~equences and the PGK-1/~Q cass~te and 10 bp from the PGK-1 promoter ~nuobotides : --518 go -507; s~e Adra ~L 1987) (Figure 1 A). One ffflh o1 the ~:
15 PCR-amplifisd samples wa-~ analyzed by ele~rophoresis through a 1.5% agaro e gel. ~3els were soaked for 30 minutes in denatunng solution (0.5 ~ NaOH, 1.5 M Nat~ the DNA tragments bl~tted for g0 minutes onto Genescreen membranes (Dupont). ;~
Blot~ed DNA was cros~-linkad to the membrane by UV light and ~ -20 hybridized in hyblidization buffer l0.5 .~a sodium phospha~e, pH
7.0, 7% SDS, 15% formamide, 1 m~ Fl)TA~ and 10 mg/mL BSA
for 3 hours at 60~C to a 132P]-labeled 850-bp DNA p~be denvsd by PCR amp~fica~ion of the short arm of pFRK90. The hybridked ~-m~mbran0 wa~ wæhed ~wice for 30 minutes with 150 m~a sodium 25 phosphat~ buffe~, pH 7.0, containing 0.1Yo SDS, on~ for 30 minu~es w~h 30 m~a sodium phosphate buffsr pH 7.0, 0.1% SDS, air dried and ~xposed to Kodalc X-Omat ~Im at -70C w~h the help of an in~nslfying s~aen. For Southem analysis of g6nomic DNA, ES cells wera grown to confluenc~ in 24-w011 plates and the DNA
30 was a~dract0d as de~ribed (Laird ~aL, 1991). DNA (10 ~9) was digest~d w~th ~amHI or ~I, electrophoressd on a û.7% agarose ge~, blotted and hybridiz~d as dsscribad abov~ with a l32p].
Iabebd 1.4 kbp 2~ dIII DNA fragment derived from phage ~12 genomic sequences located immediately upstream of the 5' end of tha targeting vector pFRK90 (Figure 1 A).
Immunoblo~ln~
Western blot analy . is Of I~kB proteins was essentially 5 performed as described (Klein ~,L. 1990a). Briefly, newborn heads or brains were homog3nized (0.1 g/mL) in immuno~
pr~cipitation buffer conta~ning 2.5 WmL aprotinin and 1 mM PMSF
and the resuiting axtracts clarified by c~ntnfugation. The catalytic gp145~5B receptors wer~ immunopracipitated with a ~B mouse 10 monoclonal antibody 18-29.2 (unpublished results), folloYved by :
incubation ~th a secondary rabbit anti-mousc 19~3 antiserum and protein A - Sepharose. The non-catalytic ~pg5~sB r~ceptor was immunopreapitated by incubafion with a rabbit polyclonal ~ ~
antiserum (#104) raised against a peptide corresponding to ~ha 13 ~ -15 carboxy-tsrminal residues ot gp9~ (Klein ~L, 1 990a).
Immunopreapitates were separated by 7.5% SDS-PAGE, blotted onto nitrocellulose fil~ers and incubated with either a cross-reactiYe rabbit polyclonal antiserum raised agains~ a peptide corraspondin~ to the 14 carboxy-terminal amino acid residues of 20 ~p140~k (Santa Cruz Biot~h., Inc.) to iden~ify gp145~ or anUs~n~m ~ 104 to id~ntity gpg5~1kB. Immunobbts ware incubat~d w~th [1251J-labebd protein A (5.6 IlC-~g, 5 ~Ci per 10 mL of Tris- :
buff6red saline~ and axposed to Kodak X-Omat film at -70C ~or the requir0d bn~th of time in the p~sence o~ int~nsifyin~ screens.
25 Hlstolo~y and blolphometrlG An31y~1s Newbsm mice (P0) from a heter~zygous ~kBlTK (~
n~ing pair wena ~rans~rdially perfused with 4~6 paraform aldehyds in PBS, decapitated, and the h~ads and bodies ,olac~d into fresh fixa~ive for 2 to 4 hours. Followin~ ~his ~hsrl post-30 hxation, tissues wers eryoprotectad in 30% sucro6~/PBS ovemightat 4C. For sectioning, heads wsre bioclced in the coronal plans and ~mbsdded in tissue ~reezing medium H-TFM (Triangle Biom~dical Scienc~s) ~t -58C. ARer allowin3 the block to warm to -26G, serial frozen sections were taken at 15 llm, thaw-mounted . . - ; . .- . . ...... ... . , . - ~. .-,; . -.- ., - . ~,. - . , .

~ ~ ~ DC;30 onto Superfrost-Plus slides (Fisher), allowed to air-dry, and -stained with cresyl violet acetate. The trigeminai ganglion and brainstem FMN were identihed and their anterior-to-postenor limits mapped. The facial motor nau-ons and trigeminal ganglion cells ~ . ~
were identified by their large size and distinct nucleus. Cells were ~ :
countad at 400X in all focal planes at 75 llm intervals by two people. ln all cas~s, variability in ~ell eounts b~ en tha two counters was less than 5/O. Areas of the two nuclei were ~ :
determin~d using a morphometric program ~SigmaScan, Jandel) at~ached to a digitizing tabl0t and drawing tube. For counts of DRG
neurons and motor neurons, whole P0 bodies were mounted in the coronal plane and serial sections were collected and stained as above. Spinal levels were determin~d by a combination of mapping the beginning and end of each individual vertabrae and DRG as wall as Ihrough characteris~ic chang~e in spinal cord shap~. Counts of DRG neurons wer~ taken from ganglia at spinal cord levals T1 1-L3. Caudal limits of the lumbar cord (L5/S1 ) were idenfffied by a dramatic dacrease in the number of motcr neurons as w811 as a reduc~ion in size of the venIral horn. Motor neurons in ~he ventral t om of the spinal cord (laminae 8 and 9) were identihed by their da~k staining cy~oplasm and pale nuclsus. Only ~hose neurons that had a visible nucl0us were counted~ Raw call counts wer~ adjusted for split nwl0i using the Ab~rcrombie (1946) correction f~tor.

~ 1 3 ~
DC~0 ~' 30 . . ~:
,.

Abl~cevlatlon~
The abbreviations used throughout this specification have the following meanings, unless othsnNise indioated in specific instances. ~ -bp base pairs BDNF brain-denv~d neurotrophic factor BSA bovine serum albumin CNS cen~ral nervous sys~em DNA dsoxyribonucleic acid DRG dwsal root ganglion EDTA ~thylenediaminetetraac0tic acid ES emblyonic stem FMN facial motornucleus HSV herpes simplex virus kbp kilo base pairs LIF leukocyte inhibitory ~actor :
NGF nerve grow~h fac~or NT neurotrophin NTP nucleetide triphosphates PAGF polyac~ylamide gel ebctrophoresis PBS pho~phate-buffered saline - ~:
PCR polymerase chain reaction PNS peripheral nervous system SDS sodium dodecyl sul~ate 2~ SEM
TK ~rosine kinase UV ultraviolet ~;;

~,.. ,. .., -,.

. .;:: : ~

~::: .: :

R~er8nc~s Abererombie, M. (1946) Estimation of nucl~ar populations ~rom microtome sections. ~m~aL~ ~L. 239.
247.
Adra, C.N., Boer, P.H., and McBurney, M.W. (1987).
Cloning and expression of the mouse pgk-1 gene and th~
nucleotide sequenca of its promoter. ~Q~Q, 65-74.
Altar, C.~, Boylan, C.B., Jackson, C., Hershenson, S., Miller, J., Wiegand, S.J., Lindsay, R.M., and tlyman, C. (1992).
Brain-derived n6urotrophic factor augments ro~ational behavior and nigrostria~al dopamine turnover ~. ~U~.~L
~, 11347-11351.
Barbacid, M. ~1993). Nerve Growth Factor: A tale of two reosp~ors. QbxgQc~, in press.
Berkameier, L.R., Winslow J.W., Kaplan, D.R., Nikolics, K., Goaddel, D.V., and Rosenthal, A. (~991). Neuro~rophin-5: A novel naurotrophic fac~or tha~ acUvates ~k and ~kB. ~ , 857-~66.
BothweU, M. (1991). Keeping track of neurotrophin receptors. ~ 6~, 915-918. ~:
Chao, M.V. (1992). Naurotrophin receptor~: a window into neuronal differ0ntiation. ~ L 583-593. . -Chusid, J.G. Correlative Neuroan~omy and FuneUonal Neurology. Los Altos: Lange Medical Publishing. 1973 D'Amico-Martel, A. and Noden, D.M. (1983).
Contributions of plaeodal and neural cres~ cells to avian cranial peripheral ganglia. ~_~, 445-46B.
DaYie~, ~ and Lumsden, A. (1984). Rebtion of target encount~r and neur~nal cell d~ath to nerve grov~h fac~or respensivaness in the developing mouse trigeminal ganglion. :
~m~ .1 24-1 37.
DeChiara, T. M., ~L. (1990) ~I~L~ ~. 78.

~i4~ DC30 DiStefano, P.S., Friedman, B., Radziejewski, C., Alexander, C., Boland, P., Schick, C.M., Lindsey, R.M. and egand, S.J. (1992). The neurotrophins BDNF, NT-3 and NGF
display distinct patterns of retrogralde a~onal transport in peripheral and central neurvns. ~ , 983-993.
Doetschman, T., Gre~g, R.G., Maeda, N." Hooper, M.L., Melton, D.W., Thompson, S., and Smithies, 0. (1987). Targeted eorrection of a mutant HPRT gene in mouse embryonic stem cells.
L~ ~Q, 576-578.
Emfors, P., Merlio, J.-P., F'ersson, H. (1992). Cells expressing mRNA for n3urotrophins and their receptors during smblyonic rat development. ~,L~4, 1140-1158.
Er~urumlu, R.S., and Killackey, H.P. (1983). Dev310pment of order in the rat trigeminal system.
, 3~5-380.
Glass, D.J., Nye, S.l 1., Hantzopoulos, P., Macchi, M.J., :
SqLIinto, S.~., Goldfarb, M., and Yancop~ulos, G.D. (1991)- ~kB ~-mediates BDNF/NT-3 dspendent survival and p~li~ration of fibroblasts lacking the low affinity NGF re~eptor. ~, 405-413. - ~ :
Hallb~8k, F., Iban~z, C.F., and Persson, H. (1991).
Evolutionary studies of the nsNa-growth factor family r~veal a novel member abundantly express9d in Xenopus ova~
~, 8~858. ~ ~ :
Hanks, 5.K., Quinn, ~M., and Huntar, T. (1988). The p~otein kina~ family: conservad featur~s and dedu~d phylogeny oftheca~alyUcdomains. ~ 1, 42-52-H~mpstead, B.L., Martin-Zanca, D., Kaplan, D.R., ~arada, L.F., and Chao, M.W. (1991~. High-affinity NGF binding requir~s 3û coexpression of the ~k proto~ncogane and the low-affinity NGF
r~ceptor. ~ ~Q, ~78-683.
Herrup, K., Di~lio, T.J. and Letsou, A. (1984). Cell lineage relationships in the development of ~he mammalian CNS. I. Ths ~acial nerve nucleu~ , 329-326.

C)C30 --~ - 33 -Hyman, C., Hofer, M., Barde, Y-A., Juhasz, M., Yancopoulos, G.D., Squinto, S.P., and Lindsay,R.M.(1991~.
BDNF is a neurotrophic factor for dopaminergic neurons of the substantia nigra. ~L~ ~Q, 230-235.
Ip N.Y.,lbanez,C.F.,Nye,S.H.,McCl~n,J.,Jones P.F., Gies, D.R., Bellusao, L., LeBeau, M.M., ~spinosa lll, R., Squinto, S.P.,Pe~son, H., and Yancnpoulos, G.D. (1992). Mammaiian neurotrophin~: structur~, chromosomal looation, tissue distribution, and receptor specificity. ~a~
1 0 ~060-3064.
Ip, N.Y., Sti~, T.N., Tapley, P., Kl~in, R., Glass D.J., Fandl, J., Gr~ene, L.A., Barbacid, M., and Yancopoulos, G.D. (1993). ~ -Similaritias and differenoes in the way neurotrophins interaGt wi~h the Trks in neuronal and non-n~uronal cells. .~lQ, 137-149. ~ ~ -Joyner, ~L., Skarnas, W.C., and RsssaRt, J. (1989). ~ -Pr~duction of a mu~ation in mouse En-2 gene by homologous rsoombination in embryonic stem cells. I!l~ ~., 163-156.
Kalcheim, C.l Barde, Y.-~, Thoenen, H. and Le Douann, N. (1987). ~Q eff~ct of a brain~erived neurotrophio ~actor on the sunrival of dovsloping dorsal root ganglion cells.
2871-2873.
Kaplan, D.R., Ma~in-Zanca, D., and Parada, L.F. (1991).
TyrDsine phosphorylation and tyrosine kinase activity of th9 ~
p~oto-oncogsne product induc~d by NGF. ~ ~Q, 158-160.
Kl~in, R., Parada, L.F., Coulier, F., and Barbaad, M.
(1989). ~dsB, a novel ty~osine protein kinase r~c~ptar expr0ssad du~in~ mou~0 neural devslopment. ~Q~. ~., 3701-3709.
Kbin, R.l Conway, D., Parada L.F., and Barbacid, M.
(199Oa~. The iEkB tyrosine pro~2in hnase g~ne codes for a second neurogenic receptor that lacks the catal~tio kinase domain. ~11 ~àl, 647-656.

h ~ DC30 Klein, R., hla~in-Zanca, D., Barbacid, M. and Parada, L.F.
(1 990b). Expression of the tyrosine hnase receptor gene ~kB is conflned to the murine embryonic and adult nervous sys~em.
~Rm~D~, 845-850.
Klein, R., Jing, S., Nanduri, V., O'Rourke, E., and Barbacid, M. (~ 991 a~. The ~k proto-oncogene encod~s a rec~ptor for nsn/e growth factor. ~11~, 189-197.
Klein, R., Nanduri, V., Jing, S., Lamballe, F.,Tapley, P., Bryant, S., Cordon-Cardo, C., Jon~s, K.R., Reichard~, L.F., and 1 O Barbacid, M. (1991 b). The ~kB tyrosine protsin kinase is a receptor for brain-derived neurotrophic fac~or and neurotrophin-3.
66, 395-403. - -~
Klein, R., Lambal~e, F., Bryant, S., and Barbacid, M.
(1 gæ). The ~kB 1yrosine protein kinase is a ~eceptor for neurotrophin-4. ~, 947-956.
Laind, P.W., Zijd~ eld, A., Linders, K., Rudnicki, M.A., Jaanisch, P~., and B0rns, A. (19913. Simplifi~d marnmalian DNA
isolation proo0dure. ~C~9~==1~, 4293.
Lamballe, F., Klein, R., and Barbacid, M. (1991). ~C, a new member of the ~k family of tyrosine protein kina~es, is a rec0ptor for neurotrophin-3. ~ i, 967-979- -Le~, K.-F., Li, E., Huber, L.J., Landis, S.C., Sharpe, AH., Chao, M.V., and Jaenisch, R. (1992). Tar~ed mutation of the ~ ~ -~n8 ~neodin0 the low affinity NaF r~ptor p5 leads to defidts - -in the peripher~l sansory nervous system. ~ ~, 737-749.
Lindsey, R.M., Tho~nen, I l., and Bar~e, Y.-A. (1985).
Plaeode ~nd neural crest-dsrived sensory n~urons ara resp~n~ive at ~a~iy d~velopm0n~al ~agas to brain-deriYed n~urotrophio factor.
~-BjQI~112~ 3~9-328.
Mansour, S.L., Thomas, K.R., and Capecchi, M.R. ~1988).
DisnJption of the prot~oncosene in~2 in mouse embry~denved str~m ~lls: a general strategy for targsting muta~ions to non-sslec~able ganes. ~ ;~, 348-352.

DC:30 lUlarsh, t l.N., Scholz, W.K., Lamballe, F., Klein, R., Nandun, V., Barbacid, M., and Palfrey, H.C. (1993). Signal transduction eYHn~s mediat~d by the BDNF receptor pg145~B in primary hippocampal pyramidal cell cultur0. L~i~, in press.
MarUn-Zanca, D., t:)skam, R., Mitra, G., Copeland, T., and Barbacid, M. (1989). Moleoular and biochemical oharac~enzation o~ th~ human h:k proto~ncogene. ~ iQL 9. 24-33-Martin-Zanca, D., Barbadd, M., and Parada, L.F. (1990).
Exprsssion of the ~.dk proto-oncogsne is restnct~d to the sensory cranial and spinal ~anglia of neurai crest origin in mouse : . -development. g~L~y. 4, 683-694.
McBurney, M.W., Sutherland, L.C., Adra, C.N., Leolair, B., Rudnidd, M.A, and Jardine, K. (1991). The mouse Pgk-1 gens promo~er contains an upstream activator sequence- ~QL~
B~Q. 5755-~761.
McKnight, S. L. (1980). ~UQL~id~ ~, 5949-Mea~dn, S.O. and Shooter, E.M. (1g92). The nen~0 ~ h factor family o~ rsceptors. ~ L~ ~, 323-331.
Merlio, J.~P., Fmfors, P., Jaber, M., ard Persson H.
(1992). Molecular olonin~ of rat ~.EkC and distfibution of cslls axpressin0 messenger P~NAs for members of ths ~k family in the ral central neNolJs sys~em. ~B~i~Q~ ~., 513-532.
bffddbl7la9, D.S., Lindber~, R.~, I lunter, T. (1991). ~kB, a neural r~cep~or protei~tyr~sine kinase: evidenoe for a ful~length and two tn~ed reeeptors. ~L~L~ l. 143-153-Opp6 nheim, R.W., Qin-Wei, Y., Pr0v0tte, D. and Yan, Q.
(1992). Brain-deriv~d n~uro~rophic factor rascues avian motor neurons ~rom cell death. ~Q ~Q, 755-757.
Rodrigu3z^Tebar, A, Dechant, a., and Barde, Y.-A. ~:
~1990). Bindin~ brain~erived neurotrophic fac~or to the nerve gro~h factor r0csptor. D~ 4, 487-492.

--~ - 36 -Rodriguez-Tebar, A., Dechant, G., Gotz, R., and Bard~, Y.-A. (19g2). Binding of neurotrophin-3 to its neuronal raceptors and interactions with nerve growth factor and brain-darived ~ ~
neurotrophic factor. ~RS2 ~1L 11, 91 7-922. ~ M ~ -S Schwarlzberg, P. L., ~L~L, (1989). ~ 2~., 799-Sendtner, M., Holtmann, B., Kolbeck, R., Thoenan, H. and E~arde, Y.-A. (1992). Brain-derived neurotrophic 1actor pr~vents the death ef motoneurons in newbom rats aner nerve sec~ion. D~
3~D, 7~7-759.
Smithies, O. ~, (1985). ~ ~Z 230.
Soppet, D., Escandon, E., Maragos, J., Middlemas, D.S., Reid, S.W., Burton, L.E., Stanton, B.R., Kaplan, D.R., I lunter, T., Nikolics, K., and Parada L.F. (1991). The neur~rophic factors ~ i brain-deriYed neurotrophic factor and n~urotrophin-3 are ligands forthe I~kBtyrosine kinase receptor. ~ i, 895-903.
Squinto, S.P., âtitt, T.N., Aldrich, T.H., Davis7 S., Bianoo, S.M., Radziejewski, G., a~ss~ D.J., Masiakowsld, P., Firth, M.E., ~aienzuela, D.M., C)iStefano, ~.S., and Yancopoulos, G.D. ~199t).
~kB encodes a functional receptor for brain-derived neurotrophic factor and neurotr~phin-3 but not nenre growth factor 885-893.
Lill9Q~ ~3~, 7~76 (19~9) Yo~el, K.S. and Davies, A.M. (1991). The duration of neuroîrophic fa~tor indepondence in ~arly sensory neuron~ is m~tched to th0 time course of targ~t fleld innerva~ion. ~l~mn ~.
819-830.
Yon Barlheld, t::.S. and Bothwell, M. (19g3).
Development of th~ mesencephalic nucleus of the trigeminal nen?e in chick embryos: Targ~t inneNation, neurotrophin ~o receptors, and cell dea~l7- ,L.a~ . 185-202.
Walton, J.N. ~
Eight Edition. Oxford: eh~ford University P~e~s. 1977.

DC3û

Wurst, W. an~ Joyner, A.L. Production of t~rgeted ambryonic stem cell clones. in ApQr~a~h, (A.L. Joyner, ed.) IRL Press, Oxfond, in press.
Yan, Q., Elliott, J. and Snider, W.D. (1992). Brain-derived 5 neurotrophio factor r0scues spinal rnotor neurons from axotomy-induced cell death. ~la~cQ~Q~ 753-755-h ~ 3 ~

Dc30 SEQUENC:E LLSTING ;

Il) GENERAL INFOR~ATION~
(i) APPL~CANT: ~lein, Rudi~er Joyner, Alexandra 3arbacid. Mariano :~
(ii) TITLE OF I~ENTION: MICE DEFICIENT IN ~TEUROTROP:-~IN RECEPTORS
i-~i) NUMBER OF SEOUENCES: 4 (i~) CORRESPONDENCE ADDRESS:
~A) ADDRESSEE: Burton Rodnev (B) STREET: P.O. Box 4000 (C1 CITY: Princeton (D~ STATE: .~ew Jerse~
(E) COUNTRY: U.S.A.
(F) ZIP: 08543-4000 (v) CO~PUTER READABLE FOR~:
(A~ MEDIUM TYPE: FloD~Y disk (B) COMPUTER: IB~ ~C com~atible (C) OPERATI~G SYSTEM: PC-DOS~MS-DOS
(D) SOFTWARE: ?aten~In Release #1.0, Versior. ~1.25 ~ : :
(vi) CURR~N~ APPLIC~TION DATA:
(A) APPLICAT~ON '`'TJM~ER:
(B j FILING D~TE:
(C) CLASSIFIC~TTON: ::
viiii ATTORNEY~AGENT TNFORMATION:
(A) NAME: ~aul. Timothv J.
(B) ~EGISTRATION NUMBER. 33.111 (C) REFERENCE/DOC~ET N~ B~R- ~C30 ~ -(ix) TELECOMMUNICATION INFORMATION: -A) TEL~P~ONE: (609) 252-5901 (B) TELEFAX: (609) 252-g526 ~ ; ~

2) I~FORMATION ~OR SEQ ID NO:l: ;
(i) SE~U~NCE C~RACTERISTICS:
(A) LENGTH: 33 base ~airs ( B ) TYPE: nucleic acid (C) STRANDEDNESS: si~le :~-~
(D) TOPOLOG;Y: linear (ii) MOLECULE TYP~: cDNA

(xij SEQUENCE DESCRIPTION: SEQ ID NO:1:
CCTTGCGGCC GCTCTTCAGA ATTTATT~AA GAG
(2) INFORMATION FOR S~Q ID NO:2:
(i) SEQUENCE C~ARACTE~ISTICS:
(A) LENGTH: 31 base ~airs ~B) TYP~: nucleic acid (C) STRANDE~NESS: si~gle (D) TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA

(xi) SEQUENCE DE5CRIPTION: S~Q ID NO:2:
GTCGCCCTCG AG~CAG~CAC CGT~GAACTT G 31 ~2) INFORMAT~ON FOR SEQ ID NO:3:
(i) SEQU~NCE CHARACT~RISTICS:
(A) LENGT~: 24 base ~airs (Bl TY~ nucleic acid (C) S~R~NDED~SS: si~gle (D) TOPOLOGY: li~ear (ii) MOLECULE TYPE: cDNA

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:3: ~ :
; GCTGGACACT GGGACTGCC~ GGCC 21 (2) INFORMATION FOR SEQ ID NO:4:
(i) SEP-uENcE C~ARACT~RISTICS:
(A~ LENG~: 22 ba~e pairs ~B) TY~E: nucleic acid . ::~
(C) ST~a~D~DNESS: singl~
(D) TOPOLOGY: 'inear (ii) MOh~CULE TYPE: c~NA `~ ~

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:4: : ~`
CTACCCGGTA GAATTCCTCG AG 22 ~ ~

Claims (7)

1. A mouse having a genome with a homozygous or heterozygous deficiency in a gene encoding a neurotrophin receptor.

2. A cell line having a genome with a homozygous or heterozygous deficiency in a gene encoding a neurotrophin receptor.

3. The mouse of Claim 1, wherein the neurotrophin receptor gene is from the trkB locus.

4. The cell line of Claim 2, wherein the neurotrophin receptor gene is from the trkB locus.

5. The cell line of Claim 2 wherein the cell line is an embryonic stem cell line.

6. The cell line of Claim 2 wherein the cell line is a pluripotent cell line.

7. The call line of Claim 2 wherein the cell line is a totipotent cell line.